Don't eliminate hand hygiene surveillance in hospitals

Yeah. Mike's correct. Hand hygiene does prevent HAI. (everyone can relax and be happy) You just can't see it or measure it and that's important. Here's why:

This past month I wrote two posts (here and here) trying to explain why you can’t link ward-level hand hygiene compliance to reduced healthcare-associated infections on the ward level. I thought the argument was pretty simple – there are so many factors in addition to hand hygiene that are important to do when preventing specific HAI (insertion checklist for CLABSI or clipping not shaving for SSI) and there are so many hand hygiene opportunities that don’t specifically impact your HAI of interest (or 30 HAI of interest) that you can’t possibly link them statistically in a study or when reporting them in your annual infection control report.

Let me be clear. I never said hand hygiene was unimportant. I said the opposite several times. Hand hygiene is critically important for infection control, particularly MDRO prevention but also HAI reduction. But for both, you can’t see the association in a graph or statistically and that’s important. Why?

In infection control, we can audit and feedback two things: process measures and outcomes.

Process measures are metrics like CLABSI checklist compliance or hand hygiene compliance. We know process measures are important for patient safety and they are easy to measure and directly remediable to education or other improvement interventions. If you see hand hygiene compliance is falling, you can educate clinicians about hand hygiene.

Outcomes measures are metrics like CLABSI rates or CAUTI rates. These are typically more difficult to measure and are delayed compared to process measures. SSI rates take a lot of person-time to track and surveillance occurs over many months for some procedures. When reviewing the literature in 2016 before giving a SHEA talk on Outcomes vs Process Measures, I found a general consensus that risk-adjusted outcomes measures are preferred to process measures because that's what the public wants to see – lower infections.

So here is why this is important. When manuscript reviewers or bloggers say ward-level hand hygiene reduces ward-level HAI, they are really saying they don’t care about hand hygiene monitoring. Specifically, they are saying – “If your intervention increases hand hygiene compliance but you can’t show me reduced HAI, I'm not interested.” We all know hand hygiene monitoring is time-consuming, costly and biased by the Hawthorne effect. So why monitor it? If hand hygiene can be linked so clearly to 30 HAI as Mike proposes, we should just go ahead and report the outcome measures (HAI rates), which will save us time and money. One less thing to do.

Let me summarize. Hand hygiene is critically important for MDRO prevention and HAI prevention. Yet, you can’t link these in clinical studies nor will you see a genuine association in your hospital’s annual report. Not seeing an association DOES NOT mean hand hygiene is unimportant.

Overcoming dogma is hard and seeing Mike’s arguments affirms my earlier trepidation. I’ll take full blame that I didn’t reference CDC and other papers like he did (although he misquoted them, to be clear). One of the things I like about blogging is that I can give you the gist without citations. But I was a bit sloppy. I’m sure I could be a better writer, but many criticisms in Mike’s post don’t hold water – perhaps I will respond in the comments or in a future post (when I am not 35,000 feet above the Pacific). For example, hand hygiene is part of the insertion bundle, but I find no mention of monthly hand hygiene compliance as being important. I'll also add that when I write these posts you are seeing my struggle to understand something. What I really appreciate about the twitter comments, both positive and negative, and Mike's post is that they make me critically determine where my ideas need more work or better explanation.

However, if we take dogma and Mike’s arguments to their logical conclusion, they suggest we can eliminate the process measure hand hygiene compliance in studies and in hospitals and replace it with HAI outcome reporting. I will say I’m surprised with this inevitable conclusion. I think that’s a huge mistake.

So there’s no misunderstanding: KEEP MONITORING HAND HYGIENE COMPLIANCE